AI Reports

Review AI-predicted customer satisfaction (CSAT), sentiment analysis, emotion detection, topic categorization, and quality insights to understand your customers' experience.

The AI tab contains 23 reports powered by AI analysis of your conversations. These reports provide insights into customer satisfaction, sentiment, emotions, and conversation topics without requiring manual surveys or tagging.

Note

AI reports use machine learning predictions based on conversation content. Only conversations with medium or high confidence scores are included in the results.

Customer Satisfaction (CSAT)

Track AI-predicted customer satisfaction scores across your organization.

CSAT Score Distribution

Shows the distribution of AI-predicted satisfaction scores on a 1–5 scale.

MetricDescription
CSAT ScoreScore from 1 (very dissatisfied) to 5 (very satisfied)
CountNumber of conversations with this score
PercentageShare of total rated conversations
Average CSATOverall average satisfaction score

Use this report to:

  • Understand overall satisfaction levels
  • Track improvements over time
  • Set benchmarks for your team

CSAT by Agent

Compares AI-predicted satisfaction scores across agents.

MetricDescription
AgentAgent name
Rated ChatsConversations with CSAT scores
Avg CSATAverage satisfaction score
Satisfied %Percentage scoring 4 or 5
Dissatisfied %Percentage scoring 1 or 2

Use this report to:

  • Compare performance across agents
  • Identify top performers
  • Spot agents who may need coaching
  • Set individual satisfaction goals
Note

A minimum number of conversations is required for a reliable comparison. Agents with very few rated conversations should be interpreted with caution.

CSAT by Channel

Compares AI-predicted satisfaction scores across communication channels.

Shows Channel, Type, Rated conversations, Average CSAT, and Satisfied percentage for each channel.

Use this report to:

  • Identify which channels deliver the best customer experience
  • Spot channels needing improvement
  • Prioritize channel investments
  • Compare satisfaction across platforms

CSAT by Business Unit

Compares AI-predicted satisfaction scores across Business Units.

Shows BU name, Rated conversations, Average CSAT, Satisfied percentage, and Dissatisfied percentage.

Use this report to:

  • Compare team satisfaction performance
  • Identify high-performing teams
  • Spot teams that may need additional support
  • Set team-level satisfaction targets

CSAT by Country

Shows AI-predicted satisfaction scores by customer country.

Displays Country, Conversations, Average CSAT, Satisfied %, Neutral %, and Dissatisfied % for each country.

Use this report to:

  • Identify low-satisfaction regions
  • Compare service quality across markets
  • Track regional satisfaction initiatives
  • Prioritize resources for underperforming regions

CSAT Trend

Tracks AI-predicted satisfaction changes day by day as a trend line.

Shows Date, Average CSAT, Rated conversations, and Satisfied percentage over time.

Use this report to:

  • Monitor satisfaction trends over time
  • Correlate changes with product updates or process changes
  • Identify the impact of training or staffing changes
  • Track progress toward satisfaction goals

Sentiment Analysis

Understand the overall mood of your customer conversations.

Sentiment Distribution

Shows the breakdown of Positive, Neutral, and Negative sentiment across conversations.

Displays each sentiment category with its count and percentage of total conversations.

Use this report to:

  • Get an overview of customer mood
  • Track sentiment trends between periods
  • Set baselines for improvement initiatives

Sentiment by Agent

Shows sentiment breakdown (Positive, Neutral, Negative) per agent with a composite Score.

MetricDescription
AgentAgent name
ChatsTotal conversations
Positive %Percentage with positive sentiment
Negative %Percentage with negative sentiment
ScoreComposite score: (Positive - Negative) / Total × 100

Use this report to:

  • Identify agents who consistently generate positive outcomes
  • Spot agents with high negative sentiment for coaching
  • Compare agent communication styles
Note

Neutral conversations pull the score toward zero. A high neutral percentage is normal and not a concern.

Sentiment by Channel

Shows sentiment breakdown for each communication channel.

Displays Channel, Type, Chats, Positive %, and Negative % per channel.

Use this report to:

  • Understand which channels produce the most positive interactions
  • Identify channels with frustration issues
  • Compare customer mood across platforms

Sentiment by Country

Shows sentiment breakdown by customer country.

Displays Country, Chats, Positive %, Neutral %, Negative %, and Score.

Use this report to:

  • Understand regional sentiment differences
  • Identify frustrated regions
  • Compare customer mood across markets
  • Prioritize regional improvements

Sentiment by Hour

Shows customer sentiment patterns by hour of day (UTC).

MetricDescription
HourHour of day (UTC)
ChatsNumber of conversations
Positive %Positive sentiment percentage
Negative %Negative sentiment percentage
ScoreComposite sentiment score (-100 to +100)

Use this report to:

  • Identify hours with higher negative sentiment
  • Correlate sentiment with staffing levels
  • Plan coverage to improve sentiment during problem hours

Sentiment Trend

Shows the proportion of Positive, Neutral, and Negative sentiment over time as a trend chart.

Displays Date, Positive %, Neutral %, Negative %, and Score for each day.

Use this report to:

  • Monitor sentiment trends over time
  • Identify unusual patterns or sentiment shifts
  • Correlate sentiment changes with events or process updates
  • Track improvement initiatives

Emotion Analysis

Detect specific emotions in your customer conversations for deeper insight beyond positive/negative sentiment.

Emotion Distribution

Shows specific emotions detected in conversations, such as Gratitude, Impatience, Frustration, and others.

Displays each emotion with its count and percentage of total conversations.

Use this report to:

  • Understand the emotional context of customer interactions
  • Identify frustration triggers
  • Recognize gratitude and positive engagement
  • Guide emotional intelligence training

Emotion by Category

Shows which emotions are detected within each topic category — revealing which issues trigger which emotional responses.

Displays Category, Total conversations, and the Dominant Emotion for each topic.

Use this report to:

  • Understand which topics trigger frustration
  • Identify gratitude-generating topics
  • Guide process improvements for high-frustration categories
  • Tailor agent training by topic and emotion

Emotion Trend

Tracks specific emotion frequency over time (Gratitude, Impatience, etc.) as a multi-line chart.

Use this report to:

  • Monitor emotional patterns over time
  • Identify periods of increased frustration
  • Track whether gratitude increases after improvements
  • Correlate emotional trends with events or changes

Category & Topic Analysis

Understand what your customers are contacting you about through AI-detected topic categories.

Category Volume

Shows conversation volume by AI-detected topic category.

ColumnDescription
CategoryAI-assigned topic
ConversationsNumber of conversations in this category
Avg CSATAverage satisfaction for this category
Negative %Percentage with negative sentiment

Use this report to:

  • Identify common inquiry types
  • Spot categories with low satisfaction
  • Prioritize training for high-volume topics
  • Track topic patterns over time

Category by Agent

Shows which AI-assigned categories each agent handles most, revealing specialization patterns.

Displays Agent, total category tags, and the Top Category for each agent.

Use this report to:

  • Identify agent specializations
  • Plan skill-based routing
  • Understand workload distribution by topic
  • Guide agent training by area of expertise
Note

Total category tags can exceed the chat count because a single conversation may be assigned multiple categories.

Category by Country

Shows what topics customers discuss, grouped by country. Different markets often have different support needs.

Results are grouped by country, showing the top 5 categories per country with count and percentage.

Use this report to:

  • Understand regional product or service differences
  • Plan localized FAQ and self-service content
  • Identify country-specific issues
  • Tailor agent training by region

Quality Insights

Advanced analysis to identify root causes of dissatisfaction and measure AI effectiveness.

Agent Sentiment Ranking

Ranks agents by overall sentiment score in a leaderboard format.

ColumnDescription
RankPosition in the leaderboard
AgentAgent name
ChatsTotal conversations
Positive %Positive sentiment percentage
Negative %Negative sentiment percentage
ScoreComposite sentiment score

Use this report to:

  • Recognize top-performing agents
  • Identify agents needing support
  • Create healthy competition
  • Track ranking changes over time
Note

A minimum of 10 conversations is required for an agent to be included in the ranking.

Dissatisfaction Drivers

Identifies which topic categories are most associated with customer dissatisfaction.

ColumnDescription
CategoryTopic category
DissatisfiedNumber of dissatisfied conversations
Avg CSATAverage satisfaction for this category
Avg WaitAverage wait time for these conversations
Avg ResponseAverage response time

Use this report to:

  • Identify improvement priorities
  • Understand correlation between wait times and dissatisfaction
  • Guide investment decisions
  • Focus training on high-dissatisfaction topics

Low CSAT Root Cause

Multi-dimensional analysis of conversations with low CSAT scores (1–2) to identify root causes of dissatisfaction.

The report includes multiple sections:

  • By Category — Which topics generate low satisfaction
  • By Emotion — Which emotions are present in dissatisfied conversations
  • By Agent — Which agents have the most low-CSAT conversations
  • By Channel — Which channels have more dissatisfied customers
  • By Hour — When low-CSAT conversations occur
  • AI Reasons — Sample AI-generated explanations for low scores

Use this report to:

  • Conduct comprehensive root cause analysis
  • Identify the most impactful improvements
  • Understand the full picture of dissatisfaction
  • Prioritize initiatives with the highest potential impact

AI Tag Analysis

Shows AI-assigned tag usage statistics across your conversations.

The report includes sections for Category Tags, Emotion Tags, and overall Totals showing how many conversations have tags, total category tag occurrences, and total emotion tag occurrences.

Use this report to:

  • Review AI categorization coverage
  • Identify the most common tags
  • Validate AI accuracy
  • Plan tag taxonomy improvements

AI Bot Containment

Measures how effectively AI agents resolve customer conversations without transferring to human agents.

MetricDescription
Total AI ChatsConversations handled by AI agents
ContainedResolved by AI without human transfer
TransferredEscalated to a human agent
Containment Rate %Percentage resolved entirely by AI

The report also includes a breakdown by channel and a daily containment trend chart.

Use this report to:

  • Evaluate AI bot efficiency and ROI
  • Identify channels where AI performs best
  • Track containment improvements over time
  • Determine what percentage of volume needs human agents
Note

A chat is considered "contained" when the AI agent resolves it without transferring to a human agent.

Sales Reports Executive Reports